INDEX
Explanations
terms related to email communication
references to emails, particularly in relation to political contexts and investigations
New Auto-Interp
Negative Logits
axis
-0.76
ESCO
-0.72
sted
-0.69
@#&
-0.68
squid
-0.67
LOS
-0.67
iculture
-0.67
INAL
-0.66
Osaka
-0.66
stad
-0.65
POSITIVE LOGITS
leaked
1.13
Emails
0.97
transcripts
0.97
dumps
0.97
declass
0.95
archived
0.93
exchanged
0.93
dump
0.93
incrim
0.92
tapes
0.92
Activations Density 0.120%